AI Governance Foundations
Oversight in Weeks
The Problem
AI is already part of how your organisation operates. People are drafting communications, preparing reports, and getting through work faster. Nobody told them to — they just found it useful. In a small team, that initiative is usually a strength.
But each person is judging what is safe, appropriate, and good enough. Without a shared understanding, those decisions add up. A gap opens between what your organisation says about quality and ethics, and what is actually happening.
A client asks how you handle their data when AI is involved. A board member wants assurance through evidence. Legislation is catching up. There is time to get ahead.
Most leaders know AI is being used but can't see where, how, or what risks it creates. A client asks how you handle their data when AI is involved — you don't have a clear answer. A board member wants assurance through evidence. Legislation is catching up: the EU AI Act is in force, UK regulators are issuing sector guidance, and stakeholders are asking "How do you govern AI?" before contracts are signed. Without documented governance, you're exposed.
Sources: KPMG/University of Melbourne — Trust in AI Global Report 2025
When people adopt AI tools without shared guidance — shadow AI — quality becomes unpredictable. Research shows 40% of AI-generated workplace content is "workslop": outputs that look finished but need significant rework. The cost is real — 4.5 hours per employee per week spent correcting AI outputs. Without shared standards, AI creates more work, not less.
Sources: BetterUp Labs/Stanford — AI Workslop Research · Zapier — AI Mistakes Survey, Jan 2026
The people most engaged with AI are often the ones asking for clear rules. Without them, there is no safe way to innovate. 44% of tech employees plan to change jobs within 12 months, many seeking organisations that take AI seriously. Only 52% of employees trust their employer to implement AI responsibly — trust is lowest where guidance is absent. At your size, losing even one good person over that frustration is felt immediately.
Sources: Korn Ferry — Tech Worker Exodus · Workday — AI Trust Gap 2025
The Solution | Proportional Oversight
Your people want to innovate responsibly with AI. That takes a shared understanding, embedded in your culture, of how your organisation uses AI well. We build it with you.
We call it Minimum Viable Governance. It gives your organisation proportional policies, processes, and oversight to innovate with AI confidently — with clear sight of risk, sized for where you are now, with room to grow. A launchpad, not a ceiling.
Your values become your boundaries, and your board gets assurance through evidence. You have it in weeks, not months.
Your team gains AI literacy — a shared language for using AI well and the confidence to innovate safely. Your clients, board, and stakeholders see an organisation that takes AI seriously. And when your team wants to try something new, you have a clear way to say yes.
When someone asks how you manage AI, you have a direct answer.
A practical, agile framework to establish AI governance built upon the concept of Minimum Viable Governance. Focuses on rapid deployment of key policies, roles and processes to manage risk without stifling innovation.
AI governance doesn't need to break the bank. We only charge for what we do — if it takes three weeks instead of four, that's the cost. A discovery call shapes the scope: we agree on the weeks and deliverables that fit your organisation, from a focused sprint through to a comprehensive deep dive. Here are two examples of what that looks like.
£1,500 per week, minimum commitment 2 weeks. Underfold undertakes client work 3 days a week, leaving 2 days for research and development. Committing to our principle of Always Evolving.
Proportional Investment
Everything you need to get governance in place and start innovating with confidence.
Rapid discovery and AI use form
AI System Inventory
AI Principles & Core Policy
Top 5 Risk Identification
Lightweight Approval Process
Brief Legislative & People Impact Report
Self-Guided AI Literacy Plan
⚡️ 3-WEEK Sprint - £4,500
A thorough review of your AI landscape with governance built around what we find.
Comprehensive discovery with stakeholder interviews
Full AI System Inventory & shadow AI amnesty
AI Principles, Policy & Optional Public Charter
Risk Register with owners and mitigations
Legislative, Regulatory & Ethics Impact Analysis
Full Approval Process with workflows
AI Portal setup
People Impact Assessment & Change Management Plan
AI Literacy Partner Recommendations & Training Plan
AI Roles & Governance Board (where needed)